The quality and cost of cancer care vary widely across hospitals and health systems. A large body of research indicates that hospitals with better outcomes tend to have large case volumes and a defined focus on cancer care. These and other data have prompted many to advocate for specialized cancer hospitals-so-called focused factories-as the best delivery system for achieving more efficient cancer care. However, reforms in the Affordable Care Act, including Accountable Care Organizations (ACO), are moving care away from this model. Although proponents argue that ACOs will improve efficiency by encouraging previously unaligned hospitals and physicians to evolve toward integrated delivery systems, others worry that the intended benefits will not accrue for patients with complex cancer diagnoses, and that ACOs will discourage referrals to hospitals with the greatest cancer expertise. To explore this issue, we propose a study that evaluates comprehensively the relationship between care delivery models and the quality, outcomes, and cost of cancer care in the United States. In the first aim, we will examine the impact of care delivery models on the quality of cancer care. Using national Medicare claims and linked SEER-Medicare data, we will identify cohorts of patients with breast, lung, prostate, and pancreatic cancer. We will then compare the quality of cancer care provided in different care delivery models based on levels of adherence with guideline recommendations and/or nationally-endorsed quality measures. We posit that cancer-focused hospitals will often have higher levels of adherence, due in part to their greater concentration of disease-specific expertise. In the second aim, we will assess the association between care delivery models and outcomes, including operative mortality, complications, hospitalizations, and long-term survival. We hypothesize that patients treated in cancer-focused hospitals will have better mortality outcomes, while greater delivery system integration will correlate with fewer hospitalizations following initial cancer therapy. In the third aim, we will examine cancer-related expenditures for these patient cohorts as they vary over time after diagnosis. We will also assess payments related to different clinical services including surgical procedures, systemic therapy, end-of-life care, and non-cancer-related treatments. We expect that cancer-focused hospitals will achieve savings around episodes of surgical care, and that patients treated in more integrated delivery systems will have lower longitudinal costs for their cancer care. Results from this study will prove invaluable to CMS policymakers and other stakeholders as they debate whether delivery system reforms deemed good for healthcare in general are also good for patients with cancer.